What a Bolt Head Marking Does Not Prove

A buyer orders 20,000 hex bolts, grade 8.8, zinc plated. The goods arrive, the heads are stamped 8.8, the certificates say 8.8, and the price was 8.8 money. Six months later a customer reports bolts shearing on a structural assembly at well under the rated load.

The bolts were not 8.8. The marking was applied by the manufacturer, and the manufacturer was not the one who would suffer the consequences.

The marking is self-declared

ISO 898-1 defines what 8.8 and 10.9 mean, and requires the property class to be marked on the head along with the manufacturer’s identification. Nothing in that standard requires an independent body to verify the marking before the bolt is sold.

So the marking is exactly as trustworthy as the plant that made it, which is why the first thing to check is whether the plant is the one that made them.

What the classes actually mean

Class	Tensile strength	Yield strength	Typical material	Hardness
8.8	800 MPa	640 MPa	Medium carbon steel, quenched and tempered	23–34 HRC
10.9	1,040 MPa	940 MPa	Alloy steel with molybdenum, quenched and tempered	32–39 HRC
12.9	1,220 MPa	1,100 MPa	Alloy steel, tighter chemistry control	39–44 HRC

The number before the dot is a tenth of the tensile strength in MPa. The number after is the ratio of yield to tensile times ten. So 8.8 means 800 MPa tensile, yield at 80 percent of that.

The substitution that actually happens

A plant that wants to shave cost does not counterfeit the marking. It makes the bolt from cheap low-carbon wire — the same Q195 or Q235 that goes into a non-structural fastener — cold-heads it, threads it, plates it, and stamps 8.8 on the head.

The result looks identical. It measures the same across the flats. It weighs the same.

What it does not have is the quench-and-temper. Low-carbon steel does not harden properly, so the bolt sits at roughly 10–15 HRC and 400–500 MPa tensile. Half the strength, at maybe 70 percent of the price.

Three tests, in increasing cost

Hardness. A portable hardness tester on the head or the shank end. Ten seconds per bolt. 8.8 should read 23–34 HRC. A low-carbon substitute reads under 20. This is the fastest screen and it catches most substitutions.

Chemistry. A handheld XRF or a spark spectrometer. For 10.9 and 12.9, look for molybdenum — a genuine alloy steel grade like 42CrMo or SCM435 has it, a substituted mild steel does not. For 8.8 the carbon content is the tell: below about 0.25 percent and it cannot reach 800 MPa after tempering.

Tensile test. Destructive, done in a lab, and the only one that settles the question. A batch sample pulled to failure on a tensometer gives you the real number. This is what a proper pre-shipment inspection does on structural fasteners, because the first two tests screen and this one decides.

The other failure nobody tests for

Hydrogen embrittlement. High-strength fasteners that are electroplated absorb hydrogen during the plating process, and if they are not baked afterwards at the right temperature for the right time, they fail in service weeks later — usually at the head-to-shank fillet, under load, with no warning.

It affects 10.9 and 12.9 far more than 8.8. The evidence is a delayed fracture with no visible corrosion and no overload. If you buy high-strength plated fasteners, the baking step is worth a written confirmation and a look at the oven log.

What to put in the purchase order

Four lines that cost nothing and change the conversation:

The property class, by standard number, not just the number — “ISO 898-1 class 8.8”
The steel grade the bolts must be made from
That the manufacturer’s head marking must identify the actual plant, not the trading company
That hardness and chemistry will be tested on a sample from the finished lot, and tensile on request

A supplier who objects to the fourth line is telling you something about the first three.

Importing structural fasteners and want the grade verified before shipment? Send us the specification — we will tell you which tests the order needs.